Abstract All potential factors within individual patients which contribute to a lack of response to a given therapy are not known, but cancer biologists have long hypothesized that distinct and disparate populations within the tumor can be selected for by therapy to outgrow and emerge as a resistant tumor. As more targeted therapies are being developed, the understanding of these subpopulations of cells within a tumor has become very important to clinical strategy. Thus, there is a need to effectively distinguish and evaluate different populations of cells in a tumor within formalin fixed tissue. These contextual evaluations are important for understanding the biology of a target, evaluating pharmacodynamic or surrogate efficacy markers, or evaluating biomarkers for a companion diagnostic approach. Immunohistochemistry (IHC) remains the most direct approach to evaluating biomarkers within tissue context, but requires a pathologist to subjectively separate the complex components of tumor tissue and the compartments of the tumor cells themselves to deliver a numerical score that is based on the staining intensity of a cell and the percentage of cells which stain. This output is considered qualitative, due to pathologist subjectivity in scoring sample regions, the inability to effectively discriminate minor differences in staining intensities for a biomarker, and the inability to deliver a dataset with sufficient sample size to overcome bias deficiencies. Furthermore, a significant amount of information content is lost in this score, eliminating the potential to identify and analyze discrete cell populations within a tumor that may be leading to refractory to therapy. In contrast, modern image analysis (IA) approaches can deliver a far more quantitative IHC score by objectively distinguishing tumor components and cellular compartments, detecting minor differences in staining intensity, and by performing this function across the whole tumor section. However, current IA approaches are designed only to report an average or thresholded intensity across the analyzed region, without reporting the cell-by-cell statistics required to identify discrete cell populations within a tumor. To answer this, we have designed Cellmap, which can analyze an IHC stained tumor tissue section which has been digitally imaged, and make multiparametric measurements about cell morphology and biomarker staining in every cell individually. This information can be reported tumor-wide, or within a specific component of the tumor, and/or within a compartment of the cell simultaneously. Cellmap can be used to make quantitative measurements which identify specific cells with specific signaling processes, and determine their location within a tumor section. This information can be used to identify and quantify discrete cell populations relevant to a disease hypothesis which are associated with a specific tumor microenvironment. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2683. doi:1538-7445.AM2012-2683